Performing inductive generalizations is critical for learning, yet there is much debate regarding the mechanisms underlying this ability. One view posits that similarity-based induction, utilizing perceptual features, may allow for increased encoding and higher memory accuracy on recognition tests. While category-based induction, utilizing semantic information, may result in limiting encoding of perceptual detail, thus resulting in decreased memory accuracy. In Experiment 1, we attempted to impair spontaneous categorization by presenting a second Working Memory load task. In Experiment 2, we attempted to impair perceptual processing by introducing a second Visual Search task. Results indicate that adult participants can rely on either mechanism when performing induction.